Crystal structure of the C3bot-RalA complex reveals a novel type of action of a bacterial exoenzyme

C3 exoenzymes from bacterial pathogens ADP-ribosylate and inactivate low-molecular-mass GTPases of the Rho subfamily. Ral, a Ras subfamily GTPase, binds the C3 exoenzymes from Clostridium botulinum and C. limosum with high affinity without being a substrate for ADP ribosylation. In the complex, the ADP-ribosyltransferase activity of C3 is blocked, while binding of NAD and NAD-glycohydrolase activity remain. Here we report the crystal structure of C3 from C. botulinum in a complex with GDP-bound RalA at 1.8 A resolution. C3 binds RalA with a helix-loop-helix motif that is adjacent to the active site. A quaternary complex with NAD suggests a mode for ADP-ribosyltransferase inhibition. Interaction of C3 with RalA occurs at a unique interface formed by the switch-II region, helix alpha3 and the P loop of the GTPase. C3-binding stabilizes the GDP-bound conformation of RalA and blocks nucleotide release. Our data indicate that C. botulinum exoenzyme C3 is a single-domain toxin with bifunctional properties targeting Rho GTPases by ADP ribosylation and Ral by a guanine nucleotide dissociation inhibitor-like effect, which blocks nucleotide exchange.     

Neosynthesis and activation of Rho by Escherichia coli cytotoxic necrotizing factor (CNF1) reverse cytopathic effects of ADP-ribosylated Rho.

Clostridium botulinum exoenzyme C3 inactivates the small GTPase Rho by ADP-ribosylation. We used a C3 fusion toxin (C2IN-C3) with high cell accessibility to study the kinetics of Rho inactivation by ADP-ribosylation. In primary cultures of rat astroglial cells and Chinese hamster ovary cells, C2IN-C3 induced the complete ADP-ribosylation of RhoA and concomitantly the disassembly of stress fibers within 3 h. Removal of C2IN-C3 from the medium caused the recovery of stress fibers and normal cell morphology within 4 h. The regeneration was preceded by the appearance of non-ADP-ribosylated RhoA. Recovery of cell morphology was blocked by the proteasome inhibitor lactacystin and by the translation inhibitors cycloheximide and puromycin, indicating that intracellular degradation of the C3 fusion toxin and the neosynthesis of Rho were required for reversal of cell morphology. Escherichia coli cytotoxic necrotizing factor CNF1, which activates Rho by deamidation of Gln(63), caused reconstitution of stress fibers and cell morphology in C2IN-C3-treated cells within 30-60 min. The effect of CNF1 was independent of RhoA neosynthesis and occurred in the presence of completely ADP-ribosylated RhoA. The data show three novel findings; 1) the cytopathic effects of ADP-ribosylation of Rho are rapidly reversed by neosynthesis of Rho, 2) CNF1-induced deamidation activates ADP-ribosylated Rho, and 3) inhibition of Rho activation but not inhibition of Rho-effector interaction is a major mechanism underlying inhibition of cellular functions of Rho by ADP-ribosylation.     

Entrapment of Rho ADP-ribosylated by Clostridium botulinum C3 exoenzyme in the Rho-guanine nucleotide dissociation inhibitor-1 complex

RhoA, -B, and -C are ADP-ribosylated by Clostridium botulinum exoenzyme C3 to induce redistribution of the actin filaments in intact cells, a finding that has led to the notion that the ADP-ribosylation blocks coupling of Rho to the downstream effectors. ADP-ribosylation, however, does not alter nucleotide binding, intrinsic, and GTPase-activating protein-stimulated GTPase activity. ADP-ribosylated Rho is even capable of activating the effector protein ROK in a recombinant system. Treatment of cells with a cell-permeable chimeric C3 toxin led to complete localization of modified Rho to the cytosolic fraction based on the complexation of ADP-ribosylated Rho with the guanine-nucleotide dissociation inhibitor-1 (GDI-1). The modified complex turned out to be resistant to phosphatidylinositol 4,5-bisphosphate- and GTPgammaS-induced release of Rho from GDI-1. Thus, ADP-ribosylation leads to entrapment of Rho in the GDI-1 complex. The increased stability of the GDI complex prevented binding of Rho to membrane-associated players of the GTPase cycle such as the activating guanine nucleotide exchange factors and effector proteins.     

A novel C3-like ADP-ribosyltransferase from Staphylococcus aureus modifying RhoE and Rnd3

Clostridium botulinum C3 is the prototype of the family of the C3-like transferases that ADP-ribosylate exclusively RhoA, -B and -C. The ADP-ribose at Asn-41 results in functional inactivation of Rho reflected by disaggregation of the actin cytoskeleton. We report on a new C3-like transferase produced by a pathogenic Staphylococcus aureus strain. The transferase designated C3(Stau) was cloned from the genomic DNA. At the amino acid level, C3(Stau) revealed an identity of 35% to C3 from C. botulinum and Clostridium limosum exoenzyme, respectively, and of 78% to EDIN from S. aureus. In addition to RhoA, which is the target of the other C3-like transferases, C3(Stau) modified RhoE and Rnd3. RhoE was ADP-ribosylated at Asn-44, which is equivalent to Asn-41 of RhoA. RhoE and Rnd3 are members of the Rho subfamily, which are deficient in intrinsic GTPase activity and possess a RhoA antagonistic cell function. The protein substrate specificity found with recombinant Rho proteins was corroborated by expression of RhoE in Xenopus laevis oocytes showing that RhoE was also modified in vivo by C3(Stau) but not by C3 from C. botulinum. The poor cell accessibility of C3(Stau) was overcome by generation of a chimeric toxin recruiting the cell entry machinery of C. botulinum C2 toxin. The chimeric C3(Stau) caused the same morphological and cytoskeletal changes as the chimeric C. botulinum C3. C3(Stau) is a new member of the family of the C3-like transferases but is also the prototype of a subfamily of RhoE/Rnd modifying transferases.     

Differential effects of Rho GTPases on axonal and dendritic development in hippocampal neurones

Formation of neurites and their differentiation into axons and dendrites requires precisely controlled changes in the cytoskeleton. While small GTPases of the Rho family appear to be involved in this regulation, it is still unclear how Rho function affects axonal and dendritic growth during development. Using hippocampal neurones at defined states of differentiation, we have dissected the function of RhoA in axonal and dendritic growth. Expression of a dominant negative RhoA variant inhibited axonal growth, whereas dendritic growth was promoted. The opposite phenotype was observed when a constitutively active RhoA variant was expressed. Inactivation of Rho by C3-catalysed ADP-ribosylation using C3 isoforms (Clostridium limosum, C3(lim) or Staphylococcus aureus, C3(stau2)), diminished axonal branching. By contrast, extracellularly applied nanomolar concentrations of C3 from C. botulinum (C3(bot)) or enzymatically dead C3(bot) significantly increased axon growth and axon branching. Taken together, axonal development requires activation of RhoA, whereas dendritic development benefits from its inactivation. However, extracellular application of enzymatically active or dead C3(bot) exclusively promotes axonal growth and branching suggesting a novel neurotrophic function of C3 that is independent from its enzymatic activity.     

ADP-ribosylation of actin isoforms by Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin

The substrate specificities of the actin-ADP-ribosylating toxins, Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin were studied by using five different preparations of actin isoforms: alpha-skeletal muscle actin, alpha-cardiac muscle actin, gizzard gamma-smooth muscle actin, spleen beta- and gamma-cytoplasmic actin, and aortic smooth muscle actin containing alpha- and gamma-smooth muscle actin isoforms. C. perfringens iota toxin ADP-ribosylated all actin isoforms tested, whereas C. botulinum C2 toxin did not modify alpha-skeletal muscle actin or alpha-cardiac muscle actin. Spleen beta/gamma-cytoplasmic actin and gizzard gamma-smooth muscle actin were substrates of C. botulinum C2 toxin. In the aortic smooth muscle actin preparation, gamma-smooth muscle actin but not alpha-smooth muscle actin was ADP-ribosylated by C. botulinum C2 toxin. The data indicate that, in contrast to C. perfringens iota toxin, C. botulinum C2 toxin ADP-ribosylates only beta/gamma-cytoplasmic and gamma-smooth muscle actin and suggest that the N-terminal region of actin isoforms define the substrate specificity for ADP-ribosylation by C. botulinum C2 toxin.     

ADP-ribosylation of skeletal muscle and non-muscle actin by Clostridium perfringens iota toxin

The enzymatically active component ia of Clostridium perfringens iota toxin ADP-ribosylated actin in human platelet cytosol and purified platelet beta/gamma-actin, in a similar way to that been reported for component I of botulinum C2 toxin. ADP-ribosylation of cytosolic and purified actin by either toxin was inhibited by 0.1 mM phalloidin indicating that monomeric G-actin but not polymerized F-actin was the toxin substrate. Perfringens iota toxin and botulinum C2 toxin were not additive in ADP-ribosylation of platelet actin. Treatment of intact chicken embryo cells with botulinum C2 toxin decreased subsequent ADP-ribosylation of actin in cell lysates by perfringens iota or botulinum C2 toxin. In contrast to botulinum C2 toxin, perfringens iota toxin ADP-ribosylated skeletal muscle alpha-actin with a potency and efficiency similar to non-muscle actin. ADP-ribosylation of purified skeletal muscle and non-muscle actin by perfringens iota toxin led to a dose-dependent impairment of the ability of actin to polymerize.     

Functional modification of a 21-kilodalton G protein when ADP-ribosylated by exoenzyme C3 of Clostridium botulinum.

Exoenzyme C3 from Clostridium botulinum types C and D specifically ADP-ribosylated a 21-kilodalton cellular protein, p21.bot. Guanyl nucleotides protected the substrate against denaturation, which implies that p21.bot is a G protein. When introduced into the interior of cells, purified exoenzyme C3 ADP-ribosylated intracellular p21.bot and changed its function. NIH 3T3, PC12, and other cells rapidly underwent temporary morphological alterations that were in certain respects similar to those seen after microinjection of cloned ras proteins. When injected into Xenopus oocytes, C3 induced migration of germinal vesicles and potentiated the cholera toxin-sensitive augmentation of germinal vesicle breakdown by progesterone, also as caused by ras proteins. Nevertheless, p21.bot was immunologically distinct from p21ras.     

Identification of rho as a substrate for botulinum toxin C3-catalyzed ADP-ribosylation

Recombinant Aplysia rho and a GTP-binding protein purified from human neutrophil membranes (G22K) were ADP-ribosylated by botulinum toxin C3 with stoichiometries of 0.8 and 0.6, respectively. Rho and G22K appeared to be different proteins since (i) rho migrated faster on polyacrylamide gels, (ii) unlike G22K, rho did not require the presence of cytosol to be ADP-ribosylated, (iii) G22K was not recognized by an anti-rho antiserum, and (iv) antibody 142-24E05 recognized G22K effectively but only poorly cross reacted with rho. ADP-ribosylation had no effect on the ability of rho to bind or hydrolyse GTP. Therefore, it appears that there are multiple botulinum toxin C3 substrates and that the toxin exerts its effects on cell function by a mechanism other than modulating the GTPase activity of rho.     

Human RhoA/RhoGDI complex expressed in yeast: GTP exchange is sufficient for translocation of RhoA to liposomes

The human small GTPase, RhoA, expressed in Saccharomyces cerevisiae is post-translationally processed and, when co-expressed with its cytosolic inhibitory protein, RhoGDI, spontaneously forms a heterodimer in vivo. The RhoA/RhoGDI complex, purified to greater than 98% at high yield from the yeast cytosolic fraction, could be stoichiometrically ADP-ribosylated by Clostridium botulinum C3 exoenzyme, contained stoichiometric GDP, and could be nucleotide exchanged fully with [3H]GDP or partially with GTP in the presence of submicromolar Mg2+. The GTP-RhoA/RhoGDI complex hydrolyzed GTP with a rate constant of 4.5 X 10(-5) s(-1), considerably slower than free RhoA. Hydrolysis followed pseudo-first-order kinetics indicating that the RhoA hydrolyzing GTP was RhoGDI associated. The constitutively active G14V-RhoA mutant expressed as a complex with RhoGDI and purified without added nucleotide also bound stoichiometric guanine nucleotide: 95% contained GDP and 5% GTP. Microinjection of the GTP-bound G14V-RhoA/RhoGDI complex (but not the GDP form) into serum-starved Swiss 3T3 cells elicited formation of stress fibers and focal adhesions. In vitro, GTP-bound-RhoA spontaneously translocated from its complex with RhoGDI to liposomes, whereas GDP-RhoA did not. These results show that GTP-triggered translocation of RhoA from RhoGDI to a membrane, where it carries out its signaling function, is an intrinsic property of the RhoA/RhoGDI complex that does not require other protein factors or membrane receptors.     

Rho protein inactivation induced apoptosis of cultured human endothelial cells

Small GTP-binding Rho GTPases regulate important signaling pathways in endothelial cells, but little is known about their role in endothelial cell apoptosis. Clostridial cytotoxins specifically inactivate GTPases by glucosylation [Clostridium difficile toxin B-10463 (TcdB-10463), C. difficile toxin B-1470 (TcdB-1470)] or ADP ribosylation (C. botulinum C3 toxin). Exposure of human umbilical cord vein endothelial cells (HUVEC) to TcdB-10463, which inhibits RhoA/Rac1/Cdc42, or to C3 toxin, which inhibits RhoA, -B, -C, resulted in apoptosis, whereas inactivation of Rac1/Cdc42 with TcdB-1470 was without effect, suggesting that Rho inhibition was responsible for endothelial apoptosis. Disruption of endothelial microfilaments as well as inhibition of p160ROCK did not induce endothelial apoptosis. Exposure to TcdB-10463 resulted in activation of caspase-9 and -3 but not caspase-8 in HUVEC. Moreover, Rho inhibition reduced expression of antiapoptotic Bcl-2 and Mcl-1 and increased proapoptotic Bid but had no effect on Bax or FLIP protein levels. Caspase-3 activity and apoptosis induced by TcdB-10463 were abolished by cAMP elevation. In summary, inhibition of Rho in endothelial cells activates caspase-9- and -3-dependent apoptosis, which can be antagonized by cAMP elevation.     

ADP-ribosylation of gelsolin-actin complexes by clostridial toxins.

ADP-ribosylation of the 1:1 (G-A) and 1:2 (G-A-A) gelsolin-actin complexes by Clostridium perfringens iota toxin and Clostridium botulinum C2 toxin was studied. Iota toxin ADP-ribosylated actin in the G-A complex from human platelets as effectively as skeletal muscle actin. The Km for NAD (4 microM) was identical for both substrates. C2 toxin ADP-ribosylated actin in the G-A complex with lower efficacy than nonmuscle actin from platelet cytosol. In the G-A-A complex both actin molecules were ADP-ribosylated by iota toxin. The G-A complex bound ADP-ribosylated actin (Ar) to form the G-A-Ar complex in which the weakly bound actin is ADP-ribosylated. Vice versa, ADP-ribosylated 1:1 gelsolin-actin complex (G-Ar) was able to bind unmodified actin to yield the G-Ar-A complex. ADP-ribosylation did not change the nucleation activity of either the G-Ar complex or the G-Ar-A complex. When monomeric actin was added to the G-A-Ar complex, polymerization of actin was delayed by about 10 min. According to a quantitative kinetic analysis, the delay of polymerization corresponded to the rate of dissociation of ADP-ribosylated actin from the G-A-Ar complex. This suggests that the nucleation activity of the G-A-A complex is inhibited by ADP-ribosylation of the weakly bound actin and that the inhibition can be removed by dissociation of ADP-ribosylated actin from the G-A-Ar complex.     

R-Ras glucosylation and transient RhoA activation determine the cytopathic effect produced by toxin B variants from toxin A-negative strains of Clostridium difficile

Clostridium difficile induces antibiotic-associated diarrhea through the production of toxin A and toxin B; the former toxin has been assumed to be responsible for the symptoms of the disease. Several toxin A-negative strains from C. difficile have recently been isolated from clinical cases and have been reported to produce toxin B variants eliciting an atypical cytopathic effect. Ultrastructural analysis indicated these toxins induce a rounding cytopathic effect and filopodia-like structures. Toxin B variants glucosylated R-Ras, and transfection with a constitutively active mutant of this GTPase protected cells against their cytopathic effect. Treatment of cells with toxin B variants induced detachment from the extracellular matrix and blockade of the epidermal growth factor-mediated phosphorylation of extracellular-regulated protein kinases, demonstrating a deleterious effect on the R-Ras-controlled avidity of integrins. Treatment with toxin B variants also induced a transient activation of RhoA probably because of inactivation of Rac1. Altogether, these data indicate that the cytopathic effect induced by toxin B variants is because of cell rounding and detachment mediated by R-Ras glucosylation, and the induction of filopodia-like structures is mediated by RhoA activation. Implications for the pathophysiology of C. difficile-induced diarrhea are discussed.     

Characterization of the ADP-ribosylation of actin by Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin

Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin belong to a novel family of actin ADP-ribosylating toxins. ADP-ribosylation of actin inhibits actin polymerization and G-actin-associated ATPase activity. The ADP-form of actin is ADP-ribosylated at a higher rate than actin with bound ATP. ADP-ribosylation of actin is reversible, a reaction, which is accompanied by reconstitution of actin ATPase activity.     

Involvement of the GTP binding protein Rho in constitutive endocytosis in Xenopus laevis oocytes

To study an endocytotic role of the GTP-binding protein RhoA in Xenopus oocytes, we have monitored changes in the surface expression of sodium pumps, the surface area of the oocyte and the uptake of the fluid-phase marker inulin. Xenopus oocytes possess intracellular sodium pumps that are continuously exchanged for surface sodium pumps by constitutive endo- and exocytosis. Injection of Clostridium botulinum C3 exoenzyme, which inactivates Rho by ADP-ribosylation, induced a redistribution of virtually all intracellular sodium pumps to the plasma membrane and increased the surface area of the oocytes. The identical effects were caused by injection of ADP-ribosylated recombinant RhoA into oocytes. The C3 exoenzyme acts by blocking constitutive endocytosis in oocytes, as determined using a mAb to the beta 1 subunit of the mouse sodium pump as a reporter molecule and oocytes expressing heterologous sodium pumps. In contrast, an increase in endocytosis and a decrease in the surface area was induced by injection of recombinant Val14-RhoA protein or Val14-rhoA cRNA. PMA stimulated sodium pump endocytosis, an effect that was blocked by a specific inhibitor of protein kinase C (Go 16) or by ADP-ribosylation of Rho by C3. Similarly, the phorbol ester-induced increase in fluid-phase endocytosis in oocytes was inhibited by Go 16, C3 transferase, or by injection of ADP-ribosylated RhoA. In contrast to C3 transferase, C. botulinum C2 transferase, which ADP-ribosylates actin, had no effect on sodium pump endocytosis or PMA-stimulated fluid- phase endocytosis. The data suggests that RhoA is an essential component of a presumably clathrin-independent endocytic pathway in Xenopus oocytes which can be regulated by protein kinase C.     

ADP-ribosylation of the bovine brain rho protein by botulinum toxin type C1

We have separated at least six GTP-binding proteins (G proteins) with Mr values between 20,000 and 25,000 from bovine brain crude membranes (Kikuchi, A., Yamashita, T., Kawata, M., Yamamoto, K., Ideda, K., Tanimoto, T., and Takai, Y. (1988) J. Biol. Chem. 263, 2897-2904). Three of these G proteins were copurified with the proteins ADP-ribosylated by botulinum toxin type C1. One G protein ADP-ribosylated by this toxin was identified to be the bovine brain rho protein (rho p20) which was purified to near homogeneity (Yamamoto, K., Kondo, J., Hishida, T., Teranishi, Y., and Takai, Y. (1988) J. Biol. Chem. 263, 9926-9932). rho p20 was ADP-ribosylated by botulinum toxin type C1 in time- and dose-dependent manners. About 0.4 mol of ADP-ribose was maximally incorporated into 1 mol of rho p20. The ADP-ribosylation of rho p20 was dependent on the presence of Mg2+. GTP enhanced the ADP-ribosylation in the presence of a low concentration (50 nM) of Mg2+ but not in the presence of a high concentration (0.5 mM) of Mg2+. The high concentration of Mg2+ fully stimulated the ADP-ribosylation even in the absence of GTP. The ADP-ribosylation of rho p20 did not affect its GTP gamma S-binding and GTPase activities. These results indicate that there are at least three G proteins ADP-ribosylated by botulinum toxin type C1 in bovine brain crude membranes and that one of them is rho p20. Two other G proteins have not yet been identified, but neither the c-ras protein, ADP-ribosylation factor for Gs, nor a G protein with a Mr of 24,000 was ADP-ribosylated by this toxin.     

Alteration of the cytoskeleton of mammalian cells cultured in vitro by Clostridium botulinum C2 toxin and C3 ADP-ribosyltransferase

The effects of Clostridium botulinum C3 ADP-ribosyltransferase and of Clostridium botulinum C2 toxin were studied on the cytoskeleton of rat hepatoma FAO and human glioma U333 cells. After treatment of these cells for 24 to 48 h with C3 (3-30 micrograms/ml), the actin microfilaments disappeared, and the intermediate filament network was found to collapse, while microtubules remained intact. Similar alterations of the cytoskeletal filaments without affecting microtubules were induced by the actin-ADP-ribosylating C2 toxin. In FAO cells, C3 caused the rounding up of cells. Concomitantly, cytosolic 22 to 24 kDa proteins were ADP-ribosylated in a guanine nucleotide-dependent manner. Rounding up of cells and ADP-ribosylation of proteins in intact cells were observed at similar concentration of the transferase. These data suggest a role of the protein substrates of C3 in the regulation of the cytoskeletal integrity.     

EhRho1, a RhoA-like GTPase of Entamoeba histolytica, is modified by clostridial glucosylating cytotoxins

Clostridial glucosylating cytotoxins inactivate mammalian Rho GTPases by mono-O glucosylation of a conserved threonine residue located in the switch 1 region of the target protein. Here we report that EhRho1, a RhoA-like GTPase from the protozoan parasite Entamoeba histolytica, is glucosylated by clostridial cytotoxins. Recombinant glutathione S-transferase-EhRho1 and EhRho1 from cell lysate of Entamoeba histolytica were glucosylated by Clostridium difficile toxin B and Clostridium novyi alpha-toxin. In contrast, Clostridium difficile toxin A, which shares the same mammalian protein substrates with toxin B, did not modify EhRho1. Change of threonine 52 of EhRho1 to alanine prevented glucosylation by toxin B from Clostridium difficile and by alpha-toxin from Clostridium novyi, which suggests that the equivalent threonine residues are glucosylated in mammalian and Entamoeba Rho GTPases. Lethal toxin from Clostridium sordellii did not glucosylate EhRho1 but labeled several other substrate proteins in lysates from Entamoeba histolytica in the presence of UDP-[14C]glucose.     

De-ADP-ribosylation actin by Clostridium perfringens iota-toxin and Clostridium botulinum C2 toxin

The reverse reaction of the ADP-ribosylation of actin by Clostridium botulinum C2 toxin and Clostridium perfringens iota-toxin was studied. In the presence of nicotinamide (30-50 mM) C2 toxin and iota-toxin decreased the radioactive labeling of [32P]ADP-ribosylated actin and catalyzed the formation of [32P]NAD. The pH optima for both reactions were 5.5-6.0. Concomitant with the removal of ADP-ribose, the ability of actin to polymerize was restored and actin ATPase activity increased. Neither ADP-ribosylation nor removal of ADP-ribose was observed after treatment of actin with EDTA, indicating that the native structure of actin is required for both reactions. ADP-ribosylation of platelet actin by C2 toxin was reversed by iota-toxin, confirming recent reports that both toxins modify the same amino acid in actin. However, C. botulinum C2 toxin was not able to cleave ADP-ribose from skeletal muscle actin which had been incorporated by iota-toxin, corroborating the different substrate specificities of both toxins.